In this study a method is proposed for detect pomegranate fruits on the tree and find the number of overall pomegranates using near camera images obtained from the stations established the groves. The pomegranate has significant red color, so a color-based method is applied for to detect the fruits on the tree. Color alone cannot provide a sufficiently robust algorithm for the detection of the pomegranate. Therefore a robust detection strategy will need to incorporate other properties of the pomegranate than just color. Pomegranate in general tends to have a round shape, where as tree branches and leaves tend to have more straight or pointed shapes. Looking for round objects can be a simple way to detect fruit. The fruit segmentation based on machine vision contains some problems (lighting and occlusion). In the proposed method these problems are also examined.